Nowadays, the web is the main collection of information resources but the retrieval of relevant and precise information from the web is a serious problem. The reason of this problem is the way web-contents are stored and formalized. With the realization of the ontology-based web, it is being tried to overcome this problem, but the ontologies may themselves suffer from heterogeneity when they are simultaneously used. Then, a need arises to identify and measure the similarity between concepts of those ontologies to overcome the said problem. Many investigators did work on this problem but still some issues need to be addressed. In this dissertation, a semi-automatic technique has been proposed for similarity identification and measurement between two web ontologies. The explicit semantics of concepts, in addition to their linguistic and taxonomic characteristics, is the key consideration and feature of the proposed technique. This proposed technique identifies all candidate pairs of similar concepts without omitting any similar-pair. The semantic relations and degree of similarity computed from new criteria, make the technique well-suited with the theme of semantic web. This technique can be used with different types of operations on ontology such as merging, mapping and aligning. We have implemented the proposed technique in Java, and validated it through different case studies. By comparing and analyzing the results of these case studies, a reasonable improvement in terms of completeness, correctness and overall quality of results, has been found.